CIO’s Guide to Oracle Database Cloud
Oracle was a late entrant to the cloud business, launching that just in 2015. Larry Ellison and Mark Hurd were reluctant as they saw the cloud as something that could cut into their revenue stream for traditional software licenses. But now they, like the vast majority of the industry, have embraced the cloud model. Here, we explain the Oracle offering.
The Oracle Database Cloud
Most IT shops now run hybrid public and private cloud infrastructures or run everything in the public cloud. There are two approaches for the database: (1) deploy virtual machines and install Oracle there, or (2) sign up to use Oracle as a service. Before Oracle’s announcement, few vendors offered option 2.
One company that does this is the mighty Amazon with their RDS service. Amazon is also where SAP runs its SAP Hana cloud database. SAP elected to use Amazon rather than stand up their own data centers. Other vendors offer other databases. Most use MySQL. Google, of course, calls their product Google SQL. Microsoft offers SQL Server. Most of these vendors also offer noSQL and other Big Data databases too, as does Oracle.
Reasons for the CIO to move their Oracle databases to the cloud are the same reasons they moved their virtual machines there: there is no upfront capital expenditure and it drives down operating costs. This is because the cloud vendor enjoys economies of scale and can pass those savings along to their customers.
Oracle is doing this too. One cost saving is on licensing fees. But CIOs still need to do careful planning. As with any cloud service, it’s important to properly size the infrastructure. Some users of AWS, in particular, have seen their monthly expenses actually exceed the cost of running all of that in-house as they were not careful in planning and over-provisioned their infrastructure.
What Does it Cost?
A company looking to drive down database operating costs by moving to the cloud will obviously want to know the answer to the question: Which cloud should we use? Security and stability are key. But which vendor costs less? Oracle, Rackspace, Amazon, or another? In short, there is no easy answer.
The reason for this is, as anyone who has looked at cloud subscription fees for virtual machines knows, there are so many variables that it’s hard to get an accurate price estimate without writing a detailed plan. A company needs a detailed architecture and inventory of IT assets because price depends on machine size, network bandwidth and utilization, storage, etc. In the past, it was simple to throw hardware at the sizing problem. But in in the cloud, there is a large price difference between, say, 7GB of memory and 25GB.
Neither Oracle nor Amazon charge by the number of records in the database, as Oracle used to do. Instead they charge a monthly or hourly rate. The rate varies by product and the size of the installation. A normal installation of Oracle Enterprise Edition is not much. Oracle Exadata machines, which are massive machines tuned for databases, cost many times more.
You can get details on Oracle Database Cloud pricing here.
Setting up a Trial
Companies looking to test Oracle’s offering can sign up for a 30-day free trial. Customers can sign up for storage, container, and backup services too as needed.
The company provides root OS and SYSDBA access, so there is no limit to what the client can do from the Oracle Linux command prompt. When the DBA creates a database instance, they are given the option of 11g or 12c. 12c would obviously be a good philosophical selection for the cloud model. This is because with 12c, where “c” could mean “cloud,” includes self-service provisioning. That is a principal characteristic that defines any cloud product.
Customers who are not ready to convert to 12c might stick with 11g.
How it Works
The basic setup procedure is go to the Oracle Database Cloud interface and sign up for a service. They offer Database as a Service, and Database Schema Service. The second option is geared more for development installations. Also, Oracle offers many more services than database, a portion of which are shown in the graphic below.
The first step to setting up an instance is to use puttygen or ssh-gen to create an SSH key used to login. That’s obviously necessary as the database is stood up with a public IP address that anyone could access. From there, the admin uses SQL Developer to create tables and add users. They add containers by sending out an authentication token and request using Curl. Then, when the database is up and running in production, they can monitor that using Oracle’s DBaaS monitor.
Here is a screenshot of the control panel where the user creates a database instance. The options include picking the size of the virtual machine, which runs from 1 to 16 CPUs and 7 to 250 GB RAM. The user can pick the Oracle Database Cloud Service, to create database instances using a wizard, or the Virtual Image, where the DBA does that manually.
Finally, having elected to use the cloud, a customer would need convert their data and upload it. Taking a shot at Oracle’s business, Amazon developed a tool to do that. Their executives rolled it out to the media. But any kind of conversion requires more than a simple tool. It requires planning in both the execution and design. Plus, the conversion provides an opportunity for the user to do much needed cleanup of the database, purge stale records, and so forth. So task your analysts with planning your conversion or engage a consulting firm.
How MiCORE Can Help
Ready to make a move to the cloud? MiCORE is here to help. As an Oracle specialized partner, our team has the expert knowledge needed to configure the cloud solution that fits your needs and seamlessly migrate your data to the cloud. Contact us today to learn more!